Data‐independent acquisition‐based SWATH‐MS for quantitative proteomics: a tutorial

Many research questions in fields such as personalized medicine, drug screens or systems biology depend on obtaining consistent and quantitatively accurate proteomics data from many samples. SWATH‐MS is a specific variant of data‐independent acquisition (DIA) methods and is emerging as a technology that combines deep proteome coverage capabilities with quantitative consistency and accuracy. In a SWATH‐MS measurement, all ionized peptides of a given sample that fall within a specified mass range are fragmented in a systematic and unbiased fashion using rather large precursor isolation windows. To analyse SWATH‐MS data, a strategy based on peptide‐centric scoring has been established, which typically requires prior knowledge about the chromatographic and mass spectrometric behaviour of peptides of interest in the form of spectral libraries and peptide query parameters. This tutorial provides guidelines on how to set up and plan a SWATH‐MS experiment, how to perform the mass spectrometric measurement and how to analyse SWATH‐MS data using peptide‐centric scoring. Furthermore, concepts on how to improve SWATH‐MS data acquisition, potential trade‐offs of parameter settings and alternative data analysis strategies are discussed.     

Systematic Evaluation of Immobilized Trypsin‐Based Fast Protein Digestion for Deep and High‐Throughput Bottom‐Up Proteomics

Immobilized trypsin (IM) has been recognized as an alternative to free trypsin (FT) for accelerating protein digestion 30 years ago. However, some questions of IM still need to be answered. How does the solid matrix of IM influence its preference for protein cleavage and how well can IM perform for deep bottom‐up proteomics compared to FT? By analyzing Escherichia coli proteome samples digested with amine or carboxyl functionalized magnetic bead–based IM (IM‐N or IM‐C) or FT, it is observed that IM‐N with the nearly neutral solid matrix, IM‐C with the negatively charged solid matrix, and FT have similar cleavage preference considering the microenvironment surrounding the cleavage sites. IM‐N (15 min) and FT (12 h) both approach 9000 protein identifications (IDs) from a mouse brain proteome. Compared to FT, IM‐N has no bias in the digestion of proteins that are involved in various biological processes, are located in different components of cells, have diverse functions, and are expressed in varying abundance. A high‐throughput bottom‐up proteomics workflow comprising IM‐N‐based rapid protein cleavage and fast CZE‐MS/MS enables the completion of protein sample preparation, CZE‐MS/MS analysis, and data analysis in only 3 h, resulting in 1000 protein IDs from the mouse brain proteome.     

Study of rat hypothalamic proteome by HPLC/ESI ion trap and HPLC/ESI‐Q‐TOF MS

The proteomic profile of hypothalamus, a key organ of CNS, is explored here by employing two widely used MS techniques, i.e. HPLC/ESI‐ion trap and HPLC/ESI‐quadrupole‐TOF MS. Strong cation exchange is used for the fractionation of peptides and protein search engine MASCOT is employed for data query. One hundred and thirty six proteins with 10 973 peptides were identified by HPLC/ESI‐ion trap MS, while 140 proteins with 32 183 peptides were characterized by HPLC/ESI‐quadrupole‐TOF MS. Among the total 198 proteins identified in both experiments, 78 proteins were common in both sets of conditions. The rest of the 120 proteins were identified distinctly in both MS strategies, i.e. 58 unique proteins were found using the quadrupole‐TOF while 62 were found with the HPLC/ESI‐ion trap. Moreover, these proteins were classified into groups based on their functions performed in the body. Results presented here identified some important signal and cellular defense proteins inevitable for survival in stressed conditions. Additionally, it is also shown that any single MS strategy is not reliable for good results due to loss of data depending on sensitivity of the instrument used.     

Targeted On‐line SPE‐LC‐MS/MS Assay for the Quantitation of 12 Apolipoproteins from Human Blood

Laborious sample pretreatment of biological samples represents the most limiting factor for the translation of targeted proteomics assays from research to clinical routine. An optimized method for the simultaneous quantitation of 12 major apolipoproteins (apos) combining on‐line SPE and fast LC‐MS/MS analysis in 6.5 min total run time was developed, reducing the manual sample pretreatment time of 3 μL serum or plasma by 60%. Within‐run and between‐day imprecisions below 10 and 15% (n = 10) and high recovery rates (94–131%) were obtained applying the high‐throughput setup. High‐quality porcine trypsin was used, which outperformed cost‐effective bovine trypsin regarding digestion efficiency. Comparisons with immunoassays and another LC‐MS/MS assay demonstrated good correlation (Pearson's R: 0.81–0.98). Further, requirements on sample quality concerning sampling, processing, and long‐term storage up to 1 year were investigated revealing significant influences of the applied sampling material and coagulant on quantitation results. Apo profiles of 1339 subjects of the LIFE‐Adult‐Study were associated with lifestyle and physiological parameters as well as establish parameters of lipid metabolism (e.g., triglycerides, cholesterol). Besides gender effects, most significant impact was seen regarding lipid‐lowering medication. In conclusion, this novel highly standardized, high‐throughput targeted proteomics assay utilizes a fast, simultaneous analysis of 12 apos from least sample amounts.     

A comparison of MS/MS‐based,stable‐isotope‐labeled,quantitation performance on ESI‐quadrupole TOF and MALDI‐TOF/TOF mass spectrometers

The peptide‐based quantitation accuracy and precision of LC‐ESI (QSTAR Elite) and LC‐MALDI (4800 MALDI TOF/TOF) were compared by analyzing identical Escherichia coli tryptic digests containing iTRAQ‐labeled peptides of defined abundances (1:1, 2.5:1, 5:1, and 10:1). Only 51.4% of QSTAR spectra were used for quantitation by ProteinPilot Software versus 66.7% of LC‐MALDI spectra. The average protein sequence coverages for LC‐ESI and LC‐MALDI were 24.0 and 18.2% (14.9 and 8.4 peptides per protein), respectively. The iTRAQ‐based expression ratios determined by ProteinPilot from the 57 467 ESI‐MS/MS and 26 085 MALDI‐MS/MS spectra were analyzed for measurement accuracy and reproducibility. When the relative abundances of peptides within a sample were increased from 1:1 to 10:1, the mean ratios calculated on both instruments differed by only 0.7–6.7% between platforms. In the 10:1 experiment, up to 64.7% of iTRAQ ratios from LC‐ESI MS/MS spectra failed S/N thresholds and were excluded from quantitation, while only 0.1% of the equivalent LC‐MALDI iTRAQ ratios were rejected. Re‐analysis of an archived LC‐MALDI sample set stored for 5 months generated 3715 MS/MS spectra for quantitation, compared with 3845 acquired originally, and the average ratios differed by only 3.1%. Overall, MS/MS‐based peptide quantitation performance of offline LC‐MALDI was comparable with on‐line LC‐ESI, which required threefold less time. However, offline LC‐MALDI allows the re‐analysis of archived HPLC‐separated samples.     

ICPLQuant – A software for non‐isobaric isotopic labeling proteomics

The main goal of many proteomics experiments is an accurate and rapid quantification and identification of regulated proteins in complex biological samples. The bottleneck in quantitative proteomics remains the availability of efficient software to evaluate and quantify the tremendous amount of mass spectral data acquired during a proteomics project. A new software suite, ICPLQuant, has been developed to accurately quantify isotope‐coded protein label (ICPL)‐labeled peptides on the MS level during LC‐MALDI and peptide mass fingerprint experiments. The tool is able to generate a list of differentially regulated peptide precursors for subsequent MS/MS experiments, minimizing time‐consuming acquisition and interpretation of MS/MS data. ICPLQuant is based on two independent units. Unit 1 performs ICPL multiplex detection and quantification and proposes peptides to be identified by MS/MS. Unit 2 combines MASCOT MS/MS protein identification with the quantitative data and produces a protein/peptide list with all the relevant information accessible for further data mining. The accuracy of quantification, selection of peptides for MS/MS‐identification and the automated output of a protein list of regulated proteins are demonstrated by the comparative analysis of four different mixtures of three proteins (Ovalbumin, Horseradish Peroxidase and Rabbit Albumin) spiked into the complex protein background of the DGPF Proteome Marker.     

Analysis of the rat hypothalamus proteome by data‐independent label‐free LC‐MS/MS

Studies of neuronal, endocrine, and metabolic disorders would be facilitated by characterization of the hypothalamus proteome. Protein extracts prepared from 16 whole rat hypothalami were measured by data‐independent label‐free nano LC‐MS/MS. Peptide features were detected, aligned, and searched against a rat Swiss‐Prot database using ProteinLynx Global Server v.2.5. The final combined dataset comprised 21 455 peptides, corresponding to 622 unique proteins, each identified by a minimum of two distinct peptides. The majority of the proteins (69%) were cytosolic, and 16% were membrane proteins. Important proteins involved in neurological and synaptic function were identified including several members of the Ras‐related protein family and proteins involved in glutamate biosynthesis.     

Machine learning reveals sex‐specific 17β‐estradiol‐responsive expression patterns in white perch (Morone americana) plasma proteins

With growing abundance and awareness of endocrine disrupting compounds (EDCs) in the environment, there is a need for accurate and reliable detection of EDC exposure. Our objective in the present study was to observe differences within and between the global plasma proteomes of sexually mature male and female white perch (Morone americana) before (Initial Control, IC) and after 17β‐estradiol (E₂) induction. Semiquantitative nanoLC‐MS/MS data were analyzed by machine learning support vector machines (SVMs) and by two‐way ANOVA. By ANOVA, the expression levels of 44, 77, and 57 proteins varied significantly by gender, treatment, and the interaction of gender and treatment, respectively. SVMs perfectly classified male and female perch IC and E₂‐induced plasma samples using the protein expression data. E₂‐induced male and female perch plasma proteomes contained significantly higher levels of the yolk precursors vitellogenin Aa and Ab (VtgAa, VtgAb), as well as latrophilin and seven transmembrane domain‐containing protein 1 (Eltd1) and kininogen 1 (Kng1). This is the first report that Eltd1 and Kng1 may be E₂‐responsive proteins in fishes and therefore may be useful indicators of estrogen induction.     

Evaluation of the Potential Risk of Advanced Peak Determination in Distorting Isobaric Labeling‐Based Single‐Shot Proteome Quantitation

The recent development and implementation of the advanced peak determination (APD) algorithm with MS instrument dramatically increased the sampling of low abundance features for MS/MS fragmentation. After in‐depth evaluation, it is found that with APD on, many chimeric spectra are acquired through co‐fragmentation of high abundance contaminants with low abundance targets, and such co‐fragmentations are largely avoided when APD is off. To evaluate whether such a co‐fragmentation could significantly distort the accuracy of the isobaric‐labeling based quantitation of the low abundance target, a single‐shot TMT experiment is performed using a two‐proteome model, whereby each TMT channel contains premixed peptides from human and a cyanobacterium with a known ratio. Unexpectedly, it is found that APD does not significantly distort TMT ratios, probably because the majority of the APD‐specific chimeric spectra are not identifiable. Nevertheless, a few examples of significant distortion of TMT ratios of low abundance peptides caused by APD is found through manual inspection, and suggests that APD should be off in a single‐shot TMT experiment to avoid the laborious and time‐costing manual inspection.     

Mass Spectrometry‐Based Analysis of Time‐Resolved Proteome Quantification

The aspect of time is essential in biological processes and thus it is important to be able to monitor signaling molecules through time. Proteins are key players in cellular signaling and they respond to many stimuli and change their expression in many time‐dependent processes. Mass spectrometry (MS) is an important tool for studying proteins, including their posttranslational modifications and their interaction partners—both in qualitative and quantitative ways. In order to distinguish the different trends over time, proteins, modification sites, and interacting proteins must be compared between different time points, and therefore relative quantification is preferred. In this review, the progress and challenges for MS‐based analysis of time‐resolved proteome dynamics are discussed. Further, aspects on model systems, technologies, sampling frequencies, and presentation of the dynamic data are discussed.     

Balancing robust quantification and identification for iTRAQ: Application of UHR‐ToF MS

iTRAQ reagents allow the simultaneous multiplex identification and quantification of a large number of proteins. Success depends on effective peptide fragmentation in order to generate both peptide sequence ions (higher mass region, 150–2200 m/z) and reporter ions (low mass region, 113–121 m/z) for protein identification and relative quantification, respectively. After collision‐induced dissociation, the key requirements to achieve a good balance between the high and low m/z ions are effective ion transmission and detection across the MS/MS mass range, since the ion transmission of the higher m/z range competes with that of the low m/z range. This study describes an analytical strategy for the implementation of iTRAQ on maXis UHR‐Qq‐ToF instruments, and discusses the impact of adjusting the MS/MS ion transmission parameters on the quality of the overall data sets. A technical discussion highlights a number of maXis‐specific parameters, their impact of quantification and identification, and their cross‐interactions.     

Comparison of detergent‐based sample preparation workflows for LTQ‐Orbitrap analysis of the Escherichia coli proteome

This work presents a comparative evaluation of several detergent‐based sample preparation workflows for the MS‐based analysis of bacterial proteomes, performed using the model organism Escherichia coli. Initially, RapiGest‐ and SDS‐based buffers were compared for their protein extraction efficiency and quality of the MS data generated. As a result, SDS performed best in terms of total protein yields and overall number of MS identifications, mainly due to a higher efficiency in extracting high molecular weight (MW) and membrane proteins, while RapiGest led to an enrichment in periplasmic and fimbrial proteins. Then, SDS extracts underwent five different MS sample preparation workflows, including: detergent removal by spin columns followed by in‐solution digestion (SC), protein precipitation followed by in‐solution digestion in ammonium bicarbonate or urea buffer, filter‐aided sample preparation (FASP), and 1DE separation followed by in‐gel digestion. On the whole, about 1000 proteins were identified upon LC‐MS/MS analysis of all preparations (>1100 with the SC workflow), with FASP producing more identified peptides and a higher mean sequence coverage. Each protocol exhibited specific behaviors in terms of MW, hydrophobicity, and subcellular localization distribution of the identified proteins; a comparative assessment of the different outputs is presented.     

Validation of a robust proteomic analysis carried out on formalin‐fixed paraffin‐embedded tissues of the pancreas obtained from mouse and human

A number of reports have recently emerged with focus on extraction of proteins from formalin‐fixed paraffin‐embedded (FFPE) tissues for MS analysis; however, reproducibility and robustness as compared to flash frozen controls is generally overlooked. The goal of this study was to identify and validate a practical and highly robust approach for the proteomics analysis of FFPE tissues. FFPE and matched frozen pancreatic tissues obtained from mice (n = 8) were analyzed using 1D‐nanoLC‐MS(MS)² following work up with commercially available kits. The chosen approach for FFPE tissues was found to be highly comparable to that of frozen. In addition, the total number of unique peptides identified between the two groups was highly similar, with 958 identified for FFPE and 1070 identified for frozen, with protein identifications that corresponded by approximately 80%. This approach was then applied to archived human FFPE pancreatic cancer specimens (n = 11) as compared to uninvolved tissues (n = 8), where 47 potential pancreatic ductal adenocarcinoma markers were identified as significantly increased, of which 28 were previously reported. Further, these proteins share strongly overlapping pathway associations to pancreatic cancer that include estrogen receptor α. Together, these data support the validation of an approach for the proteomic analysis of FFPE tissues that is straightforward and highly robust, which can also be effectively applied toward translational studies of disease.     

Proteomic profiles of human lung adeno and squamous cell carcinoma using super‐SILAC and label‐free quantification approaches

Nonsmall cell lung cancer (NSCLC) accounts for 85% of lung cancers, and is subdivided into two major histological subtypes: adenocarcinoma (ADC) and squamous cell carcinoma (SCC). There is an unmet need to further subdivide NSCLC according to distinctive molecular features that may be associated with responsiveness to therapies. Four primary tumor‐derived xenograft proteomes (two‐each ADC and SCC) were quantitatively compared by using a super‐SILAC labeling approach together with ultrahigh‐resolution MS. Proteins highly differentially expressed in the two subtypes were identified, including 30 that were validated in an independent cohort of 12 NSCLC primary tumor‐derived xenograft tumors whose proteomes were quantified by an alternative, label‐free shotgun MS methodology. The 30‐protein signature contains metabolism enzymes including phosphoglycerate dehydrogenase, which is more highly expressed in SCC, as well as a comprehensive set of cytokeratins and other components of the epithelial barrier, which is therefore distinctly different between ADC and SCC. These results demonstrate the utility of the super‐SILAC method for the characterization of primary tissues, and compatibility with datasets derived from different MS‐based platforms. The validation of proteome signatures of NSCLC subtypes supports the further development and application of MS‐based quantitative proteomics as a basis for precision classifications and treatments of tumors. All MS data have been deposited in the ProteomeXchange with identifier PXD000438 ( http://proteomecentral.proteomexchange.org/dataset/PXD000438 ).     

Characterization of the Proteome of Theobroma cacao Beans by Nano‐UHPLC‐ESI MS/MS

Cocoa seed storage proteins play an important role in flavour development as aroma precursors are formed from their degradation during fermentation. Major proteins in the beans of Theobroma cacao are the storage proteins belonging to the vicilin and albumin classes. Although both these classes of proteins have been extensively characterized, there is still limited information on the expression and abundance of other proteins present in cocoa beans. This work is the first attempt to characterize the whole cocoa bean proteome by nano‐UHPLC‐ESI MS/MS analysis using tryptic digests of cocoa bean protein extracts. The results of this analysis show that >1000 proteins could be identified using a species‐specific Theobroma cacao database. The majority of the identified proteins were involved with metabolism and energy. Additionally, a significant number of the identified proteins were linked to protein synthesis and processing. Several proteins were also involved with plant response to stress conditions and defence. Albumin and vicilin storage proteins showed the highest intensity values among all detected proteins, although only seven entries were identified as storage proteins. A comparison of MS/MS data searches carried out against larger non‐specific databases confirmed that using a species‐specific database can increase the number of identified proteins, and at the same time reduce the number of false positives. The results of this work will be useful in developing tools that can allow the comparison of the proteomic profile of cocoa beans from different genotypes and geographic origins. Data are available via ProteomeXchange with identifier PXD005586.     

In‐depth characterization of the tomato fruit pericarp proteome

Since the genome of Solanum lycopersicum L. was published in 2012, some studies have explored its proteome although with a limited depth. In this work, we present an extended characterization of the proteome of the tomato pericarp at its ripe red stage. Fractionation of tryptic peptides generated from pericarp proteins by off‐line high‐pH reverse‐phase phase chromatography in combination with LC‐MS/MS analysis on a Fisher Scientific Q Exactive and a Sciex Triple‐TOF 6600 resulted in the identification of 8588 proteins with a 1% FDR both at the peptide and protein levels. Proteins were mapped through GO and KEGG databases and a large number of the identified proteins were associated with cytoplasmic organelles and metabolic pathways categories. These results constitute one of the most extensive proteome datasets of tomato so far and provide an experimental confirmation of the existence of a high number of theoretically predicted proteins. All MS data are available in the ProteomeXchange repository with the dataset identifiers PXD004947 and PXD004932.     

Plasma and liver proteomic analysis of 3Z‐3‐[(1H‐pyrrol‐2‐yl)‐methylidene]‐1‐(1‐piperidinylmethyl)‐1,3‐2H‐indol‐2‐one‐induced hepatotoxicity in Wistar rats

3Z‐3‐[(1H‐pyrrol‐2‐yl)‐methylidene]‐1‐(1‐piperidinylmethyl)‐1,3‐2H‐indol‐2‐one (Z24), a synthetic anti‐angiogenic compound, inhibits the growth and metastasis of certain tumors. Previous works have shown that Z24 induces hepatotoxicity in rodents. We examined the hepatotoxic mechanism of Z24 at the protein level and looked for potential biomarkers. We used 2‐DE and MALDI‐TOF/TOF MS to analyze alternatively expressed proteins in rat liver and plasma after Z24 administration. We also examined apoptosis in rat liver and measured levels of intramitochondrial ROS and NAD(P)H redox in liver cells. We found that 22 nonredundant proteins in the liver and 11 in the plasma were differentially expressed. These proteins were involved in several important metabolic pathways, including carbohydrate, lipid, amino acid, and energy metabolism, biotransformation, apoptosis, etc. Apoptosis in rat liver was confirmed with the terminal deoxynucleotidyl transferase dUTP‐nick end labeling assay. In mitochondria, Z24 increased the ROS and decreased the NAD(P)H levels. Thus, inhibition of carbohydrate aerobic oxidation, fatty acid β‐oxidation, and oxidative phosphorylation is a potential mechanism of Z24‐induced hepatotoxicity, resulting in mitochondrial dysfunction and apoptosis‐mediated cell death. In addition, fetub protein and argininosuccinate synthase in plasma may be potential biomarkers of Z24‐induced hepatotoxicity.     

Glucosyltransferase‐dependent and ‐independent effects of TcdB on the proteome of HEp‐2 cells

Toxin B (TcdB) of the nosocomial pathogen C. difficile has been reported to exhibit a glucosyltransferase‐dependent and ‐independent effect on treated HEp‐2 cells at toxin concentration above 0.3 nM. In order to investigate and further characterize both effects epithelial cells were treated with wild type TcdB and glucosyltransferase‐deficient TcdB_NXN and their proteomes were analyzed by LC‐MS. Triplex SILAC labeling was used for quantification. Identification of 5212 and quantification of 4712 protein groups was achieved. Out of these 257 were affected by TcdB treatment, 92 by TcdB_NXN treatment and 49 by both. TcdB mainly led to changes in proteins that are related to “GTPase mediated signaling” and the “cytoskeleton” while “chromatin” and “cell cycle” related proteins were altered by both, TcdB and TcdB_NXN. The obtained dataset of HEp‐2 cell proteome helps us to better understand glucosyltransferase‐dependent and ‐independent mechanisms of TcdB and TcdB_NXN, particularly those involved in pyknotic cell death. All proteomics data have been deposited in the ProteomeXchange with the dataset identifier PXD006658 ( https://proteomecentral.proteomexchange.org/dataset/PXD006658 ).